1. Field of the Invention
The present invention relates to an image processing apparatus and method, and more particularly, to an image processing apparatus and method by which it is possible to suppress degradation of a encoding efficiency by locally controlling filtering during encoding or decoding.
2. Description of the Related Art
Recently, apparatuses complying with a standard such as the MPEG (Moving Picture Experts Group) are commercially available in both transmitting information from a broadcasting center or the like and receiving information in customer premises. Such apparatuses digitally handle and compress image information using an orthogonal conversion such as a discrete cosine transform and a motion compensation based on characteristic redundancy of the image information in order to transmit and accumulate information with an excellent efficiency.
Particularly, the MPEG2 (ISO(International Organization for Standardization)/IEC(International Electrotechnical Commission) 13818-2) defined as a general image encoding scheme is currently used in a wide variety fields such as both professional and consumer applications, both interlaced and the progressive scanning images, and is a standard which encompasses both standard and high resolution images. According to the MPEG2 compression scheme, for example, a bit rate of 4 to 8 Mbps is allocated to a standard resolution interlaced image of 720×480 pixels, and a bit rate of 18 to 22 Mbps is allocated to a high resolution interlaced image of 1920×1088 pixels so that a high compression rate and an excellent image quality can be realized.
While the MPEG2 is mainly targeted to a high quality image encoding suitable for a broadcasting, it fails to provide an encoding scheme having a bit rate lower than that of the MPEG1, i.e., with a higher compression rate. As mobile phones are widely used, it is anticipated that demands on such an encoding scheme increases in the future. Accordingly, the MPEG4 encoding scheme has been standardized. As the image encoding scheme, such a standard has been internationally approved as ISO/IEC 14496-2 in December 1998.
Furthermore, recently, for the purpose of encoding a TV conference image, H.26L (ITU-T (ITU Telecommunication Standardization Sector) Q6/16 VCEG (Video encoding Experts Group)) is being standardized. According to H.26L, it is known that a large amount of computations are necessary to perform the encoding and the decoding in comparison with existing encoding schemes such as MPEG2 or MPEG4, but higher encoding efficiency can be realized. In addition, currently, as one of the MPEG4 activities, a standardization for realizing higher encoding efficiency is being developed as a Joint Model of Enhanced-Compression Video encoding on the basis of H.26L by including additional functionalities that are not supported by H.26L. As a schedule of the standardization, an international standard based on the names of H.264 and MPEG4 Part 10 (AVC (Advanced Video Coding)) has been approved in March 2003.
Recently, as a next-generation video encoding technology, an adaptive loop filter (ALF) is being reviewed (e.g., refer to Yi-Jen Chiu and L. Xu, “Adaptive (Wiener) Filter for Video Compression,” ITU-T SG16 Contribution, C437, Geneva, April 2008). This adaptive filter can alleviate block distortion or quantization distortion that is difficult to handle using a deblock filter by performing an optimal filtering for each frame.
However, generally, since the image locally has various characteristics, the optimal filter coefficient is locally different. In the method disclosed in the Yi-Jen Chiu and L. Xu, “Adaptive (Wiener) Filter for Video Compression,” ITU-T SG16 Contribution, C437, Geneva, April 2008, since the same filter coefficient is applied to all pixels in a single frame, image quality may be improved in the entire frame but may be degraded locally.
In this regard, a method of omitting the filtering in the area where image quality is locally degraded has been proposed (e.g., refer to Takeshi. Chujoh, et al., “Block-based Adaptive Loop Filter,” ITU-T SG16 Q6 VCEG Contribution, AI18, Germany, July, 2008, and T. Chujoh, N. Wada, and G. Yasuda, “Quadtree-based Adaptive Loop Filter, ” ITU-T SG16 Q6 VCEG Contribution, VCEG-AK22(r1), Japan, April, 2009). In this method, the image encoding apparatus controls whether or not the filtering is performed for each control block by matching with a plurality of control blocks compactly arranged on the area of the image. The image encoding apparatus establishes flag information for each block and performs the adaptive filtering based on the flag information. Similarly, the image decoding apparatus performs the adaptive filtering based on the flag information.
However, a method of performing the image encoding or decoding for each slice by dividing a single frame into a plurality of slices (multi-slice) has been proposed. Takeshi. Chujoh, et al., “Block-based Adaptive Loop Filter,” ITU-T SG16 Q6 VCEG Contribution, AI18, Germany, July, 2008, and T. Chujoh, N. Wada, and G. Yasuda, “Quadtree-based Adaptive Loop Filter, ” ITU-T SG16 Q6 VCEG Contribution, VCEG-AK22(r1), Japan, April, 2009 describe a method of establishing blocks for a single entire frame and creating and transmitting flag information for all of the blocks but fail to describe how to process the flag information in such a multi-slice case and how to create and use the flag information.
Takeshi. Chujoh, et al., “Block-based Adaptive Loop Filter,” ITU-T SG16 Q6 VCEG Contribution, AI18, Germany, July, 2008, and T. Chujoh, N. Wada, and G. Yasuda, “Quadtree-based Adaptive Loop Filter, ” ITU-T SG16 Q6 VCEG Contribution, VCEG-AK22(r1), Japan, April, 2009 just describe that the image encoding apparatus creates the flag information of all blocks within the frame for a single slice within the frame. That is, even in the case of the multi-slice, the image encoding apparatus is to create the flag information of all blocks within the frame for each slice.
However, the flag information of the blocks of the area other than the processing-target slice is not necessary. The created flag information is encoded together with the image data and included in the image compression information. That is, according to the methods disclosed in Takeshi. Chujoh, et al., “Block-based Adaptive Loop Filter,” ITU-T SG16 Q6 VCEG Contribution, AI18, Germany, July, 2008, and T. Chujoh, N. Wada, and G. Yasuda, “Quadtree-based Adaptive Loop Filter, ” ITU-T SG16 Q6 VCEG Contribution, VCEG-AK22(r1), Japan, April, 2009, when the multi-slice is applied, the image compression information may unnecessarily increase and degrade encoding efficiency.